(Compaq Portable I) Advice: Bad Caps replaced, PSU still bad

[BitsOfDestiny27]

Hey everyone,

I hope someone can help me with this. So I identified caps C37 and C39 on the PSU as being shorted since they blew (presumably on the first time I turned it on, since I may have heard a very small pop; makes sense the "pop" being very low since these caps are 2.2uF and quite small). I replaced them with some tantalum caps of the same rating (2.2uF 25V), and redid the solder a second time under the suspicion that the soldering may not have been perfect (first time soldering a component in).

In any case, I tested the PSU again, with no load except the MFM hard disk I plugged into one of the two molex connectors, and the HDD green light went on for a bit, with the multimeter showing up to 3V that plummet shortly after. I did the same test with my multimeter attached to one of the molex connectors, testing the 5V and 12V rails. One device was attached to the other molex as a "dummy" load, first the MFM and then the FDD. Same deal.

I have reason to believe the capacitors I installed were done so fine, albeit please feel free to judge (but be nice) my soldering job. I made sure no soldering bridges or flux residue was left on the PCB, but still, I could be wrong. Moreover, unlike the first couple of times I did these same tests (before getting to the PSU), I got higher voltage the first fraction of a second, whereas those first times before replacing the capacitors, my multimeter read less than 1V before plummeting.

Anyways, I have been watching EEVblog's vid on his diagnosis on a compaq portable PSU, having done tests such as the one he does at around the 10:00 timestamp, but I haven't properly tested for any open/shorted diodes or anything like that. I kindly ask for some guidance, or perhaps someone has had issues like myself with the PSU. Thank you so much!

P.S. The pics I am attaching are the most recent (on my "fix" of solder for one of the caps, bc on the top view of the board it seemed like the little holes where the legs go it needed to be filled as it looked like an empty bucket, if that makes sense--the other reason is I had excess solder that looked too clunky).

 

[NeXT]

The problem with older tantalum is that I have seen it fail short while in service and not just when given power.
Case in point was a Seagate ST251 last month that the PSU came up and the drive started but not 15 seconds in the tantalum promptly failed short and shut the power supply down.
While I can't say you had a tant short out in your drive I'd suspect there's still a tantalum somewhere that may of failed in the same manner. Likewise if you were seeing 3v instead of 12 or 5 that tells me either it's poorly regulating or something was heavily loading the rail down before the power supply cut out.

 

[BitsOfDestiny27]

The problem with older tantalum is that I have seen it fail short while in service and not just when given power.
Case in point was a Seagate ST251 last month that the PSU came up and the drive started but not 15 seconds in the tantalum promptly failed short and shut the power supply down.
While I can't say you had a tant short out in your drive I'd suspect there's still a tantalum somewhere that may of failed in the same manner. Likewise if you were seeing 3v instead of 12 or 5 that tells me either it's poorly regulating or something was heavily loading the rail down before the power supply cut out.

Hey! Thank you for your response. I could try the CRT as load to confirm that the issue lies in either the HDD and/or the FDD. If it fails too, then yeah the PSU has some other subtle problem. I didn't see any more visibly blown tant caps on the PSU, so I'm gonna have to look around some more. I know there's pico PSU cards, but honestly I really want to try figuring this out before buying replacements. Chances are the system board or the video card have a bad tant(s) which will only be discoverable if this PSU gets to be fixed ;-;

 

[NeXT]

Tantalums if the PSU doesn't have enough umpf will short and never visually indicate as such.